This project makes use of newly available 3D culture methods to grow patient-derived xenografts (PDXs) in vitro. 3D cultures are amenable to high throughput screens, where we can expose multiple 3D culture models representative of aggressive variant prostate cancer to promising new therapies or drugs that target biologically important pathways. Single agents that show promise will be selected for matrix combination screens, and winning combinations will be validated in PDX models in vivo. Single agent screens have identified a small number of compounds with efficacy in specific models. Initial validation experiments have generally confirmed these results, and combination screens are in initial phases.